Part one delved into the theory and justification for modern Content Management Systems (CMS). Let’s get down to the business of actually selecting a suitable CMS! First, some pitfalls to avoid:
• Don’t choose one just because your friend likes it. This is about your needs and your company, not theirs.
• Don’t choose one based on cultural attachment.
• Don’t choose one just because your IT department says you have to. Consider the entire scope of your content.
A better CMS design process follows this workflow:
- Consider your INPUTS
a. Content Strategy (substance, structure, workflow, governance)
b. Editorial Resources (the editorial team’s ongoing time commitment)
c. Technical Resources (hardware infrastructure, technical team’s ongoing time commitment)
- Use them to inform the DESIGN PROCESS
a. Content Modeling
b. Task Analysis
- Implement and govern your resulting CMS SELECTION & CUSTOMIZATION PLAN
Content Strategy and Editorial Resources
The root of your decisions rests here—the weightiest inputs, as it were. Be prepared to take into consideration:
• Editorial Strategy (i.e., “product development for content”): editorial calendar, workflows, style guide, etc.
• Content Analysis: content inventory, gap analysis, taxonomy, migration plan, etc.
• Copywriting/Information Architecture: content templates (a.k.a. page tables), copy decks, annotated wireframes, etc.
Your model should be designed to describe the content therein: types, models, and relationships. The content model semantically organizes the content, and thus can be seen as a subset of information architecture, as long as you don’t confuse it with the sitemap—that’s for top-down navigation. Ultimately, the content model is up to you. There’s no one-size-fits all; it’s more dependent on your style and judgment. That said, these guideposts can be of great assistance.
Semantics vs. Granularity
You’re basically pitting what does the content represent? against how much detail is necessary? To borrow Kahn’s example, suppose your company runs conferences and needs a content model. What are you representing? Events, presentations, speakers, and attendees. How much detail? This is usually the harder question. You could model multiple conference tracks and schedules, you could limit yourself to just presentations and speakers…But wait, there’s more! You also need to consider how content types relate to one another. Are presentations tied to tracks or events? Are these one-to-one relationships or one-to-many? How many data fields/elements does each speaker require in your scheme? These questions all boil down to classification and taxonomy. If your content is large or diverse enough to merit a CMS to handle it, it’s very important to agree on these answers before you commit to a design.
The Model Proper
With your deliverables worked out, you can sketch a working content model. The point is to clarify content types by possible associated elements, and also to demonstrate the relationships between content types visually. In the conference company example, you might want to model an event as a discrete number of tracks, each composed of a few presentations, each presented by one speaker. Thanks to semantic richness, you can flexibly choose several display options:
• Users could navigate from a speaker’s biography to past presentations
• Presentations could be shown in the context of a daily schedule, with before/after/simultaneous events
• Intelligent search could return results by speaker, event, topic, etc.
• Users could assemble a personalized schedule and export it in the form of their choice
It’s more important to have a pragmatic design than a flawless one. As your business evolves, you should be prepared for incremental changes to the model. Any number of constraints influence the design at this stage—know that you’ll probably have to compromise, or defer certain types of functionality until you’re better able to accommodate them. The only wrong choice would be to ignore these eventualities.
Model in hand, you’re now ready to address the tasks involved: editing, publishing, and renewing the content you’ve specified. Remember, though, you don’t have all the time in the world. Task Analysis is designed to acquaint your model with reality by assessing the true cost of content and features, rectifying assumptions, and allotting the proper time periods for editors and techies.
You can use this four-step guide to evaluate the editorial process:
1. Brainstorm key tasks based on your content model, editorial calendar, and content inventory
2. Sketch workflow diagrams by task (sound familiar?)
3. Sketch wireframes for interfaces
4. Estimate the editorial time required for each task
With the conference company example, you might lay out the editorial task analysis like this:
Process: find event
Decision: does the track exist? Yes: choose track. No: add track.
Decision: does a speaker exist? Yes: choose speaker. No: add speaker.
Process: add presentation and corresponding fields
Process: publish presentation.
Pretty simple so far. The usual implementation would include about five separate screens, taking up to 20 minutes to complete. When sketching wireframes, consider design elements such as auto-complete or show/hide. By repeating this process for each editorial task, you can then prioritize scope against the actual timeframe—here’s where informed adjustments can be made, where before you could only assume. You can also address the order of tasks, which may surprise you. Still, better to be surprised now than after time or money has been wasted.
The Big Finish: CMS Selection and Customization!
Now that all the planning is done, you can ruthlessly trim down the seemingly infinite list of CMS tools, eliminating any that don’t provide your functional requirements. Similarly, you can fast-track the customization discussion because you already have a concrete idea of what you want. Of course, consultation with your tech team, vendors, or online communities may still be in order. How much customization do you need, can your current resources cover it, how long until the customizations are in place? Don’t let any of these questions go unanswered, and don’t be afraid to scale back on some elements of your model to arrive at a suitable answer. After all, you’re still pre-commitment. Get estimates, and get estimates for possible workarounds!
From here, it’s the usual song and dance with corporate project plans. The system is awesome and you’re satisfied with how it came about, but that doesn’t mean you can just walk away. Good news, though: all that workflow diagramming pays off, because it’s perfectly applicable to applied content strategy for a working business. You’re playing for keeps now, but you’ve certainly practiced enough to get a feel for your capacity. Think of the difference: a well-strategized, custom CMS is a (better) universe away from the sloppy, disorganized content “plan” must companies use. This is how it’s done!